Siliques or pods from Brassica plants release their seeds through a process called fruit dehiscence. A silique consists of two carpels joined margin to margin. The suture between the margins forms a thick rib, called replum. As pod maturity approaches, the two valves separate progressively from the replum, along designated lines of weakness in the pod, eventually resulting in the shattering of the seeds that were attached to the replum. The dehiscence zone defines the exact location of the valve dissociation.
Shedding of seed (also referred to as “seed shatter” or “pod shatter”) by mature pods before or during crop harvest is a universal phenomenon with crops that develop dry dehiscent fruits. Premature seed shatter results in a reduced seed recovery, which represents a problem in crops that are grown primarily for the seeds, such as oil-producing Brassica plants, particularly oilseed rape. Another problem related to premature seed shattering is an increase in volunteer growth in the subsequent crop year. In oilseed rape, pod shatter-related yield losses are on average 20% (Child et al., 1998, J Exp Bot 49: 829-838), but can reach up to 50%, depending on the weather conditions (MacLeod, 1981, Harvesting in Oilseed Rape, pp. 107-120, Cambridge Agricultural Publishing, Cambridge).
Current commercial oilseed rape varieties are extremely susceptible to shattering. There is little variation for resistance to shattering within existing breeding programs of B. napus but resistant lines have been found within the diploid parents of B. napus (B. oleracea and B. rapa) as well as within other members of the Brassica genus, notably B. juncea, B. carinata and B. nigra. Kadkol et al. (1986, Aust. J. Botany 34 (5): 595-601) report increased resistance towards shattering in certain accessions of B. campestris that was associated with the absence of a separation layer in the region of attachment of the siliqua valves to the replum. Prakash and Chopra (1988, Plant breeding 101: 167-168) describe the introgression of resistance to shattering in Brassica napus from Brassica juncea through non-homologous recombination. Spence et al. (1996, J of Microscopy 181: 195-203) describe that some lines of Brassica juncea show a reduced tendency to shatter as compared to Brassica napus lines. Morgan et al., 1998 (Fields Crop Research 58, 153-165) describe genetic variation for pod shatter resistance among lines of oilseed rape developed from synthetic B. napus and conclude that lines which required much energy to open their pods appeared to have increased vascularisation in the dehiscence zone and to have reduced cell wall degradation within the dehiscence zone. They further found a significant negative correlation between the length of the pod beak and the force needed to cause pod shattering. Child and Huttly (1999, Proc 10th Int. Rapeseed Congress) describe variation in pod maturation in an irradiation-induced mutant B. napus and a population of its parent cultivar, Jet Neuf, wherein the most resistant wild-type and mutant plants showed much lignification of groups of cells throughout the dehiscence zone and wherein vascular traces situated close to the inner edge of the dehiscence zone in the mutant were described to help to secure the valves. Child et al. (2003, J Exp Botany 54 (389): 1919-1930) further describe the association between increased pod shatter resistance and changes in the vascular structure in pods of a resynthesized Brassica napus line. However, the traditional methods for breeding have been unsuccessful in introducing shatter resistance into rape cultivars, without interference with other desirable traits such as early flowering, maturity and blackleg resistance (Prakash and Chopra, 1990, Genetical Research 56: 1-2).
Several genes, which promote or inhibit pod dehiscence, have been identified in Arabidopsis thaliana through mutant analysis: Combined mutants in both SHATTERPROOF1 (SHP1; initially referred to as AGL1) and SHATTERPROOF2 (SHP2; initially referred to as AGL5) result in indehiscent siliques (i.e. siliques which remain closed upon maturity in Arabidopsis thaliana) (Liljegren et al., 2000, Nature 404, 766-770). Similarly, mutants in the INDEHISCENT gene (referred to as IND1) in Arabidopsis thaliana (Liljegren et al., 2004, Cell 116: 843-853; PCT publication WO 01/79517), as well as in ALCATRAZ (referred to as ALC; Rajani et al. 2001, Current Biology 11, 1914-1922) interfered with pod dehiscence leading to pod shatter resistance. Constitutive expression of FRUITFUL (FUL), a repressor of SHP and IND, in Arabidopsis thaliana also resulted in indehiscent siliques (Ferrandiz et al., 2000, Science, 289, 436-438). These transcription factors are believed to form a non-linear transcriptional network that controls valve margin identity and pod shatter. Liljegren et al. (2004, Cell 116: 843-853) further describe that IND, an atypical basic helix-loop-helix (bHLH) gene, directs the differentiation of the valve margin into the separation and lignified layers in Arabidopsis thaliana. The layer of lignified cells adjacent to the separation layer along with the endocarp b layer (a single lignified cell layer in each valve) produce a spring-like tension within the drying fruit that contributes to its opening. Lignification of the valve endodocarp b layer requires the activities of IND, SHP, ALC, and FUL, a MADS-domain transcription factor that is expressed throughout the valves (Liljegren et al., 2004, supra; Mandel and Yanofsky, 1995, Plant Cell 7, 1763-1771). FUL and REPLUMLESS (RPL), a homeodomain transcription factor that is expressed in the replum (Roeder et al., 2003, Curr Biol 13, 1630-1635), have been found to set the boundaries of the genes that confer valve margin identity (Gu et al., 1998, Development 125, 1509-1517; Ferrandiz et al., 2000, Science, 289, 436-438; Roeder et al., 2003, supra). Finally, FILAMENTOUS FLOWER (FIL) and YABBY3 (YAB3), two YABBY-family transcription factors (Sawa et al., 1999, Genes Dev 13, 1079-1088; Siegfried et al., 1999, Development 126, 4117-4128), and JAGGED (JAG), a C2H2 zinc-finger transcription factor (Dinneny et al., 2004, Development 131, 1101-1110; Ohno et al., 2004, Development 131, 1111-1122), were identified to redundantly contribute to proper valve and valve margin development by promoting the expression of FUL and SHP in a region-specific manner (Dinneny et al., 2005, Development 132, 4687-4696). Genes for a number of hydrolytic enzymes, such as endopolygalacturonases, which play a role, during pod dehiscence, in the programmed breakdown of the dehiscence zone in pods from Brassica plants have also been identified (see e.g. WO 97/13865; Petersen et al., Plant. Mol. Biol., 1996, 31:517-527).
Liljegren et al. (2004, Cell 116: 843-853) describe five mutant alleles of Arabidopsis IND. The lignified cells in the dehiscence zone are either absent or present in plants comprising these mutant alleles depending on the severity of the mutations (severe ind mutants do not contain lignified cells in the region corresponding to the inner part of the valve margin in wild-type plants), but in all cases the silique is indehiscent. Wu et al. (2006), Planta 224, 971-979) describe a sixth mutant allele of Arabidopsis IND. Plants comprising this mutant allele show no lignified cells at the junctions of the valve margin and the replum, contain fewer cells in a region of seven layers of cells, which appeared to encompass the commonly known dehiscence zone and replum border in wild-type plants, and exhibit incomplete cytokinesis in this layer.
US 2005/0120417 and US 2007/0006336 describe the identification and isolation of two IND1 orthologs from Brassica napus. 
WO99/00503, WO01/79517 and WO0159122 describe down-regulation of the expression of the Arabidopsis ALC, IND, AGL1 and AGL5 genes and orthologs thereof using gene-silencing techniques (such as antisense suppression or cosuppression) and mutagenesis.
Vancanneyt et al., 2002 (XIII International Conference on Arabidopsis Research, Sevilla, Spain Jun. 28-Jul. 2; 2002) reported that the expression of FUL from A. thaliana under control of a CaMV 35S promoter in oilseed rape resulted in a number of pod shatter resistant transformants. Pods of such pod shatter resistant lines had no dehiscence zone, and opening of the pods could only be achieved by random fracture of the valves by applying considerable pressure.
Vancanneyt et al., 2002 (XIII International Conference on Arabidopsis Research, Sevilla, Spain Jun. 28-Jul. 2; 2002) also reported that silencing of the IND gene in Arabidopsis thaliana using so-called dsRNA silencing techniques resulted in almost complete pod shatter resistance. Ninety-eight percent of the transgenic Arabidopsis lines developed siliques, which did not open along the valve suture, and could only be opened by applying considerable pressure to the valves.
It is important to realize that while seed shattering constitutes an important problem in oilseed rape culture, which may be solved by developing pod shatter resistant lines, ultimately, separation of the seeds from the pods is still required. In normal agricultural practice this is achieved by treshing of the pods by a combine harvester. Treshing of the pods by a combine harvester must be complete and must cause minimum damage to the seeds thus released. However, as pod strength increases, the more severe action required to tresh them causes an unacceptable level of damage to the seed. The pods of pod shatter resistant Brassicaceae plants should thus not be so strong that they cannot be treshed in a combine harvester (Bruce et al. 2001, J. Agric. Engng Res. 80, 343-350).
WO 2004/113542 describes that moderate dsRNA gene silencing of genes involved in the development of the dehiscence zone and valve margins of pods in Brassicaceae plants allows the isolation of transgenic lines with increased pod shatter resistance and reduced seed shattering, the pods of which however may still be opened along the dehiscence zone by applying limited physical forces.
WO09/068313 (claiming priority of European patent application EP 07023052) discloses Brassica plants comprising at least two IND genes, in particular Brassica napus plants, characterized in that they comprise three full knock-out mutant IND alleles in their genome and wherein the pod shatter resistance of the plants is significantly increased compared to the pod shatter resistance of a plant not comprising mutant IND alleles, but wherein the plant preferably maintains an agronomically relevant treshability of the pods.
The inventions described hereinafter in the different embodiments, examples and claims provide further improved methods and means for modulating dehiscence properties in dehiscent seed plants. More specifically, the present invention describes further improved methods and means for reducing seed shattering, or delaying seed shattering until after harvest, in plants such as Brassicaceae plants, particularly Brassicaceae plants grown for seed production, while maintaining at the same time an agronomically relevant treshability of the pods. In particular, the present application discloses Brassica plants comprising at least two IND genes, in particular Brassica napus plants, characterized in that they comprise two partial knock-out mutant IND alleles in their genome or two partial and two full knock-out mutant IND alleles and wherein the pod shatter resistance of the plants is significantly increased compared to the pod shatter resistance of a plant not comprising mutant IND alleles, but wherein the plant preferably maintains an agronomically relevant treshability of the pods. Also provided are methods and means to increase the yield, particularly grain and seed yield. The yield increase phenotype may be separate from the reduced or delayed seed shatter phenotype.